CN105631052A - Artificial intelligence based retrieval method and artificial intelligence based retrieval device - Google Patents
Artificial intelligence based retrieval method and artificial intelligence based retrieval device Download PDFInfo
- Publication number
- CN105631052A CN105631052A CN201610115420.3A CN201610115420A CN105631052A CN 105631052 A CN105631052 A CN 105631052A CN 201610115420 A CN201610115420 A CN 201610115420A CN 105631052 A CN105631052 A CN 105631052A
- Authority
- CN
- China
- Prior art keywords
- user
- retrieval
- term
- result
- feedback
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/904—Browsing; Visualisation therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an artificial intelligence based retrieval method and an artificial intelligence based retrieval device. The artificial intelligence based retrieval method includes: acquiring a retrieval word; calculating to obtain a retrieval result corresponding to the retrieval word according to an MDP model; displaying the retrieval result; acquiring feedbacks of the retrieval result from a user to make it convenient for recalculating to obtain a retrieval result according to the MDP module and displaying a retrieval result obtained by recalculation. By the artificial intelligence based retrieval method, effectiveness in interaction with the user is achieved, user demands can be better met, and user experience is promoted.
Description
Technical field
The present invention relates to Internet technical field, particularly relate to a kind of search method based on artificial intelligence and device.
Background technology
Artificial intelligence (ArtificialIntelligence, AI) is research, the new technological sciences developing the theory for simulating, extend and expand the intelligence of people, method, application system. Artificial intelligence is a branch of computer science, the essence of intelligence is understood in its attempt, and produce a kind of intelligent machine can made a response in the way of human intelligence is similar newly, the research in this field comprises intelligent robot, speech recognition, pattern recognition, natural language processing and expert systems etc.
Search engine is as current internet one important application, it is intended to by the information display of user search to user. Existing retrieval system, is only taking the keyword of user's offer as index, recalls the result of a series of static state. But, in practice, the demand of user often shows as the process of a serializing, and when the demand of user has had horizontal or when longitudinally expanding, existing retrieval system cannot be formed real mutual with user.
Summary of the invention
One of technical problem that the present invention is intended to solve in correlation technique at least to a certain extent.
For this reason, it is an object of the present invention to propose a kind of search method based on artificial intelligence, the method can more effective carry out better meeting consumers' demand alternately with user, promotes Consumer's Experience.
Another object of the present invention is to propose a kind of indexing unit based on artificial intelligence.
For achieving the above object, the search method based on artificial intelligence that first aspect present invention embodiment proposes, comprising: obtain term; According to MDP model, calculate the result for retrieval corresponding with described term; Show described result for retrieval; Obtain user to the feedback of described result for retrieval, obtain result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
The search method based on artificial intelligence that first aspect present invention embodiment proposes, by obtaining the feedback of user, can carry out repeatedly mutual with user, thus more effective carry out alternately with user, in addition, by adopting MDP model to calculate result for retrieval, it is possible to better meet consumers' demand, promote Consumer's Experience.
For achieving the above object, the indexing unit based on artificial intelligence that second aspect present invention embodiment proposes, comprising: acquisition module, for obtaining term; Calculate module, for according to MDP model, calculating the result for retrieval corresponding with described term; Display module, for showing described result for retrieval; Feedback module, for obtaining user to the feedback of described result for retrieval, obtains result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
The indexing unit based on artificial intelligence that second aspect present invention embodiment proposes, by obtaining the feedback of user, can carry out repeatedly mutual with user, thus more effective carry out alternately with user, in addition, by adopting MDP model to calculate result for retrieval, it is possible to better meet consumers' demand, promote Consumer's Experience.
The aspect that the present invention adds and advantage will part provide in the following description, and part will become obvious from the following description, or be recognized by the practice of the present invention.
Accompanying drawing explanation
The present invention above-mentioned and/or additional aspect and advantage will become obviously with it should be readily understood that wherein from the following description of the accompanying drawings of embodiments:
Fig. 1 is the schematic flow sheet of the search method based on artificial intelligence that one embodiment of the invention proposes;
Fig. 2 is the schematic flow sheet of the search method based on artificial intelligence that another embodiment of the present invention proposes;
Fig. 3 is the structural representation of the indexing unit based on artificial intelligence that another embodiment of the present invention proposes.
Embodiment
Being described below in detail embodiments of the invention, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar module or has module that is identical or similar functions from start to finish. It is exemplary below by the embodiment being described with reference to the drawings, only for explaining the present invention, and limitation of the present invention can not be interpreted as. On the contrary, embodiments of the invention comprise all changes within the scope of the spirit and intension falling into attached claim book, amendment and etc. jljl.
Fig. 1 is the schematic flow sheet of the search method based on artificial intelligence that one embodiment of the invention proposes. See Fig. 1, the method comprises:
S11: obtain term.
Wherein, time initial, user can input term and initiate retrieval, thus search engine can receive the term of user's input.
User can with form input terms such as text, voice, pictures.
S12: according to MDP model, calculates the result for retrieval corresponding with described term.
In the present embodiment, based on " strengthening study " (ReinforcementLearning) technology in machine learning techniques as, search problem is regarded a markov decision process (MarkovDecisionProcess, MDP).
MDP model tlv triple is expressed as follows: state (state), behavior (action), feedback (reward).
MDP to be solved behavior A, and wherein a kind of mode that solves selects behavior when making financial value maximum, is formulated as:
A=argmaxA{ Q (S, A) } formula (1)
Above-mentioned formulation: solve A when making the value of Q maximum.
Wherein, Q is revenue function, and is the function about S and A, and S is state (state), A is behavior (action).
The functional form of Q is determined by R, such as, determines the functional form of Q by solving R=Q (S, A). Concrete, Q can also represent and is: Q (S, A)=sum (r0+r1+r2+...), is the financial value of each step, utilizes financial value, is learnt to Q (S, A) by temporaldifference.
Wherein, R is feedback (reward). When time initial, user does not also produce feedback, feedback is empty, and value can represent with 0.
The mode of the above-mentioned A of solving is the strategy adopting financial value maximum, it is commonly referred to Greedy, what can also adopt other solves mode, such as, adopts Explore&Exploit mode, the feature of Explore&Exploit mode is, it not select currently best every time, but have certain probability to go to select time excellent or uncertain OK strategy, comprise ��-greedy, softmax, sampling.
In the present embodiment, when introducing MDP model when retrieving, the above-mentioned tlv triple of MDP model is distinguished specifically:
S=state=term+context, wherein, term+context is corresponding with current state, for term, when different, term can be different, such as, different according to state, term can respectively: the term (when such as user clicks the term of recommendation) of user recommended by the term (when such as initial user initiates search) of user's input, search engine, user switch after term while again initiating retrieval (such as user do not meet result for retrieval). In addition, context is as comprised the recent behavior of user, browse record etc.
A=action=result for retrieval=display (Query, R), wherein, R is the web results of usual form, and for directly meeting consumers' demand, Query is the term that user recommended by search engine, excites for water conservancy diversion. The web results of usual form is such as in the webpage link of PC end display, or the result of the card form in mobile terminal display. The A corresponding with term can determine according to formula (1).
User's behavior that R=reward=user produces according to the result for retrieval of display, such as comprise: user clicks buying behavior (such as, showing the purchase information of commodity in result for retrieval, user have purchased commodity based on this), user click certain result enter corresponding web page after the residence time (referred to as length when clicking), the residence time of user in whole retrieving (referred to as length during search), user is to the term etc. after the click of result for retrieval (term of web results and/or recommendation), the switching of user's input.
Therefore, adopt above-mentioned S, A, the R in retrieving and above-mentioned formula (1) that the A corresponding with term can be solved, also it is exactly result for retrieval corresponding to term.
S13: show described result for retrieval.
After search engine gets result for retrieval, it is possible to be sent to client terminal and show.
S14: obtain user to the feedback of described result for retrieval, obtain result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
Common retrieving is exactly a reciprocal process, and in the present embodiment, user can carry out repeatedly mutual with search engine, the feedback adjustment result for retrieval of user when search engine is repeatedly mutual.
Such as, see Fig. 2, the mutual retrieval flow of many wheels can comprise:
S21: user initiates retrieval.
Such as, user inputs initial term, and after clicking search button, it is possible to initiate retrieval.
S22: search engine calculates result for retrieval according to MDP model, and shows result for retrieval.
Result for retrieval action=display (Query, R) represents.
The result for retrieval A (action) corresponding with current term can adopt formula (1) to calculate. When initially not feeding back, will feed back and manage as vacancy.
S23: the first feedback receiving user.
The first feedback clicks certain web results for user, represents in the drawings for reward (click).
S24: recalculate result for retrieval, and show result for retrieval.
Result for retrieval action=display (Query, R) represents.
The result for retrieval A (action) corresponding with current term can adopt formula (1) to calculate, and feedback wherein adopts the first above-mentioned feedback.
S25: the 2nd kind of feedback receiving user.
2nd kind of feedback clicks the term of search engine recommendation for user, represents for QueryR (clickquery) in the drawings.
S26: recalculate result for retrieval, and show result for retrieval.
Result for retrieval action=display (Query, R) represents.
The result for retrieval A (action) corresponding with current term can adopt formula (1) to calculate, and feedback wherein adopts above-mentioned the 2nd kind to feed back.
User can perform S27 or S28 afterwards.
S27: the third feedback receiving user.
The third feedback switches term for user, represents in the drawings for QueryR (search).
Such as, user is after getting result for retrieval, it is possible to not webpage clicking result, does not also click the term of recommendation, but re-enters new term.
Afterwards, search engine can recalculate result for retrieval, and shows result for retrieval.
Result for retrieval action=display (Query, R) represents.
The result for retrieval A (action) corresponding with current term can adopt formula (1) to calculate, and feedback wherein adopts the third above-mentioned feedback.
S28: terminate.
Such as, user is after getting result for retrieval, it is possible to do not carry out follow-up retrieval, terminates retrieval flow.
Above-mentioned be fed back to example with three kinds, it will be appreciated that in actual retrieval process, be not limited to user and carried out above-mentioned three kinds of feedbacks, user can carry out above-mentioned a kind of any two or carry out the feedback of other kinds. In addition, interaction times is also not limited to three times, it is also possible to carry out the mutual of other number of times, can adopt identical or different types of feedback during different times mutual.
In the present embodiment, by obtaining the feedback of user, it is possible to carry out repeatedly mutual with user, thus more effective carry out alternately with user, in addition, by adopting MDP model to calculate result for retrieval, it is possible to better meet consumers' demand, lifting Consumer's Experience. Further, by using the one of length during search as feedback, owing to the determination of behavior is relevant with feedback, therefore can using length during search as optimization aim so that user stops the longer time in a session of retrieval. By comprising the term of web results and recommendation at result for retrieval, it is possible to consider meeting to excite with water conservancy diversion as a whole. By above-mentioned feedback, term and web results (query-item) can be set up, two kinds of terms (query-query), web results and term (item-query) etc. are many to staggered guiding with meet, and can effectively make the closed loop that search is ecological. Being excited by water conservancy diversion and according to feedback adjustment result for retrieval, it is possible to the demand of user has been carried out horizontal and vertical clarification, what more focus on is the process of the whole retrieval of user, but not the recalling of single term.
Fig. 3 is the structural representation of the indexing unit based on artificial intelligence that another embodiment of the present invention proposes. See Fig. 3, this device 30 comprises: acquisition module 31, calculating module 32, display module 33 and feedback module 34.
Acquisition module 31, for obtaining term.
Wherein, time initial, user can input term and initiate retrieval, thus search engine can receive the term of user's input.
User can with form input terms such as text, voice, pictures.
Calculate module 32, for according to MDP model, calculating the result for retrieval corresponding with described term.
In the present embodiment, based on " strengthening study " (ReinforcementLearning) technology in machine learning techniques as, search problem is regarded a markov decision process (MarkovDecisionProcess, MDP).
MDP model tlv triple is expressed as follows: state (state), behavior (action), feedback (reward).
MDP to be solved behavior A, and wherein a kind of mode that solves selects behavior when making financial value maximum, represents with formula (1).
In some embodiments, the parameter of the described MDP model that described calculating module 32 adopts comprises:
State, adopts term and context to represent;
Behavior, adopts result for retrieval to represent;
Feedback, adopts user to the feedback representation of described result for retrieval.
In some embodiments, described term comprises:
The term of user's initial input, recommends the term of user, or the term after user's switching.
In some embodiments, described result for retrieval comprises:
Web results, and, recommend the term of user.
It is one or more that described feedback comprises in following item:
User is to the click of described web results;
User is to the click of the described term recommending user;
Term after the switching of user's input;
The click buying behavior of user;
During click long;
During search long.
Concrete computation process see the explanation in embodiment of the method, can illustrate at this no longer in detail.
Display module 33, for showing described result for retrieval.
After search engine gets result for retrieval, it is possible to be sent to client terminal and show.
Feedback module 34, for obtaining user to the feedback of described result for retrieval, obtains result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
Common retrieving is exactly a reciprocal process, and in the present embodiment, user can carry out repeatedly mutual with search engine, the feedback adjustment result for retrieval of user when search engine is repeatedly mutual.
The mutual retrieval flow of many wheels see a Fig. 2, can illustrate at this no longer in detail.
It should be appreciated that the present embodiment is corresponding with above-mentioned embodiment of the method, particular content see the associated description in embodiment of the method, can illustrate at this no longer in detail.
In the present embodiment, by obtaining the feedback of user, it is possible to carry out repeatedly mutual with user, thus more effective carry out alternately with user, in addition, by adopting MDP model to calculate result for retrieval, it is possible to better meet consumers' demand, lifting Consumer's Experience.
It should be noted that, in describing the invention, term " first ", " the 2nd " etc. are only for describing object, and can not be interpreted as instruction or hint relative importance. In addition, in describing the invention, unless otherwise explanation, the implication of " multiple " refers at least two.
Describe and can be understood in schema or in this any process otherwise described or method, represent and comprise one or more for realizing the module of the code of the performed instruction of the step of specific logical function or process, fragment or part, and the scope of the preferred embodiment of the present invention comprises other realization, wherein can not according to order that is shown or that discuss, comprise according to involved function by the mode while of basic or by contrary order, carrying out n-back test, this should be understood by embodiments of the invention person of ordinary skill in the field.
It is to be understood that each several part of the present invention can realize with hardware, software, firmware or their combination. In the above-described embodiment, multiple step or method can realize with the software stored in memory and perform by suitable instruction execution system or firmware. Such as, if realized with hardware, the same with in another enforcement mode, can realize with the arbitrary item in following technology well known in the art or their combination: the discrete logic with the logic gates for data signal being realized logic function, there is the application specific integrated circuit of suitable combinational logic gating circuit, programmable gate array (PGA), field-programmable gate array (FPGA) etc.
Those skilled in the art are appreciated that realizing all or part of step that above-described embodiment method carries is can be completed by the hardware that program carrys out instruction relevant, described program can be stored in a kind of computer-readable recording medium, this program perform time, step comprising embodiment of the method one or a combination set of.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, it is also possible to is that the independent physics of each unit exists, it is also possible to two or more unit are integrated in a module. Above-mentioned integrated module both can adopt the form of hardware to realize, it is also possible to adopts the form of software function module to realize. If described integrated module realize using the form of software function module and as independent production marketing or when using, it is also possible to be stored in a computer read/write memory medium.
The above-mentioned storage media mentioned can be read-only storage, disk or CD etc.
In the description of this specification sheets, at least one embodiment that the description of reference term " embodiment ", " some embodiments ", " example ", " concrete example " or " some examples " etc. means to be contained in the present invention in conjunction with concrete feature, structure, material or feature that this embodiment or example describe or example. In this manual, the schematic representation of above-mentioned term is not necessarily referred to identical embodiment or example. And, the concrete feature of description, structure, material or feature can combine in an appropriate manner in any one or more embodiment or example.
Although above it has been shown and described that embodiments of the invention, it is understandable that, above-described embodiment is exemplary, can not be interpreted as limitation of the present invention, and above-described embodiment can be changed, revises, replace and modification by the those of ordinary skill of this area within the scope of the invention.
Claims (10)
1. the search method based on artificial intelligence, it is characterised in that, comprising:
Obtain term;
According to MDP model, calculate the result for retrieval corresponding with described term;
Show described result for retrieval;
Obtain user to the feedback of described result for retrieval, obtain result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
2. method according to claim 1, it is characterised in that, the parameter of described MDP model comprises:
State, adopts term and context to represent;
Behavior, adopts result for retrieval to represent;
Feedback, adopts user to the feedback representation of described result for retrieval.
3. method according to claim 1 and 2, it is characterised in that, described term comprises:
The term of user's initial input, recommends the term of user, or the term after user's switching.
4. method according to claim 1 and 2, it is characterised in that, described result for retrieval comprises:
Web results, and, recommend the term of user.
5. method according to claim 4, it is characterised in that, it is one or more that described feedback comprises in following item:
User is to the click of described web results;
User is to the click of the described term recommending user;
Term after the switching of user's input;
The click buying behavior of user;
During click long;
During search long.
6. the indexing unit based on artificial intelligence, it is characterised in that, comprising:
Acquisition module, for obtaining term;
Calculate module, for according to MDP model, calculating the result for retrieval corresponding with described term;
Display module, for showing described result for retrieval;
Feedback module, for obtaining user to the feedback of described result for retrieval, obtains result for retrieval and the result for retrieval obtained is recalculated in display to recalculate according to MDP model.
7. device according to claim 6, it is characterised in that, the parameter of the described MDP model that described calculating module adopts comprises:
State, adopts term and context to represent;
Behavior, adopts result for retrieval to represent;
Feedback, adopts user to the feedback representation of described result for retrieval.
8. device according to claim 6 or 7, it is characterised in that, described term comprises:
The term of user's initial input, recommends the term of user, or the term after user's switching.
9. device according to claim 6 or 7, it is characterised in that, described result for retrieval comprises:
Web results, and, recommend the term of user.
10. device according to claim 9, it is characterised in that, it is one or more that described feedback comprises in following item:
User is to the click of described web results;
User is to the click of the described term recommending user;
Term after the switching of user's input;
The click buying behavior of user;
During click long;
During search long.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610115420.3A CN105631052A (en) | 2016-03-01 | 2016-03-01 | Artificial intelligence based retrieval method and artificial intelligence based retrieval device |
| KR1020160132955A KR20170102411A (en) | 2016-03-01 | 2016-10-13 | Searching method and device based on artificial intelligence |
| JP2016220983A JP6333342B2 (en) | 2016-03-01 | 2016-11-11 | Search method and apparatus based on artificial intelligence |
| US15/392,017 US20170255879A1 (en) | 2016-03-01 | 2016-12-28 | Searching method and device based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610115420.3A CN105631052A (en) | 2016-03-01 | 2016-03-01 | Artificial intelligence based retrieval method and artificial intelligence based retrieval device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105631052A true CN105631052A (en) | 2016-06-01 |
Family
ID=56045984
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610115420.3A Pending CN105631052A (en) | 2016-03-01 | 2016-03-01 | Artificial intelligence based retrieval method and artificial intelligence based retrieval device |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20170255879A1 (en) |
| JP (1) | JP6333342B2 (en) |
| KR (1) | KR20170102411A (en) |
| CN (1) | CN105631052A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108345941A (en) * | 2017-01-23 | 2018-07-31 | 阿里巴巴集团控股有限公司 | A kind of parameter regulation means and device |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3058227B1 (en) * | 2016-10-27 | 2018-11-02 | Thales | FLEXW MULTIFUNCAL RADAR, IN PARTICULAR FOR AUTOMOBILE |
| US11347751B2 (en) * | 2016-12-07 | 2022-05-31 | MyFitnessPal, Inc. | System and method for associating user-entered text to database entries |
| JP6881150B2 (en) | 2017-08-16 | 2021-06-02 | 住友電気工業株式会社 | Control devices, control methods, and computer programs |
| US11157488B2 (en) * | 2017-12-13 | 2021-10-26 | Google Llc | Reinforcement learning techniques to improve searching and/or to conserve computational and network resources |
| US20210319098A1 (en) * | 2018-12-31 | 2021-10-14 | Intel Corporation | Securing systems employing artificial intelligence |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101261634A (en) * | 2008-04-11 | 2008-09-10 | 哈尔滨工业大学深圳研究生院 | Studying method and system based on increment Q-Learning |
| CN101751437A (en) * | 2008-12-17 | 2010-06-23 | 中国科学院自动化研究所 | Web active retrieval system based on reinforcement learning |
| CN102456018A (en) * | 2010-10-18 | 2012-05-16 | 腾讯科技（深圳）有限公司 | Interactive search method and device |
| CN104035958A (en) * | 2014-04-14 | 2014-09-10 | 百度在线网络技术（北京）有限公司 | Search method and search engine |
| CN104331459A (en) * | 2014-10-31 | 2015-02-04 | 百度在线网络技术（北京）有限公司 | Online learning-based network resource recommendation method and device |
| CN104573015A (en) * | 2015-01-12 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Information searching method and device |
| CN104573019A (en) * | 2015-01-12 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Information searching method and device |
| CN105183850A (en) * | 2015-09-07 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Information querying method and device based on artificial intelligence |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000305932A (en) * | 1999-04-20 | 2000-11-02 | Nippon Telegr & Teleph Corp <Ntt> | Document retrieving method accompanied by presentation of related word, document retrieving device and recording medium recording program |
| JP4770868B2 (en) * | 2008-04-21 | 2011-09-14 | ソニー株式会社 | Information providing apparatus, information providing method, and computer program |
| JP2010033442A (en) * | 2008-07-30 | 2010-02-12 | Ntt Docomo Inc | Search system evaluation device, and search system evaluating method |
| JP2010282402A (en) * | 2009-06-04 | 2010-12-16 | Kansai Electric Power Co Inc:The | Retrieval system |
| JP5451673B2 (en) * | 2011-03-28 | 2014-03-26 | ヤフー株式会社 | Search ranking generation apparatus and method |
| JP5620604B1 (en) * | 2014-05-12 | 2014-11-05 | 株式会社ワイワイワイネット | Ranking system for search results on the net |
-
2016
- 2016-03-01 CN CN201610115420.3A patent/CN105631052A/en active Pending
- 2016-10-13 KR KR1020160132955A patent/KR20170102411A/en not_active Application Discontinuation
- 2016-11-11 JP JP2016220983A patent/JP6333342B2/en active Active
- 2016-12-28 US US15/392,017 patent/US20170255879A1/en not_active Abandoned
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101261634A (en) * | 2008-04-11 | 2008-09-10 | 哈尔滨工业大学深圳研究生院 | Studying method and system based on increment Q-Learning |
| CN101751437A (en) * | 2008-12-17 | 2010-06-23 | 中国科学院自动化研究所 | Web active retrieval system based on reinforcement learning |
| CN102456018A (en) * | 2010-10-18 | 2012-05-16 | 腾讯科技（深圳）有限公司 | Interactive search method and device |
| CN104035958A (en) * | 2014-04-14 | 2014-09-10 | 百度在线网络技术（北京）有限公司 | Search method and search engine |
| CN104331459A (en) * | 2014-10-31 | 2015-02-04 | 百度在线网络技术（北京）有限公司 | Online learning-based network resource recommendation method and device |
| CN104573015A (en) * | 2015-01-12 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Information searching method and device |
| CN104573019A (en) * | 2015-01-12 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Information searching method and device |
| CN105183850A (en) * | 2015-09-07 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Information querying method and device based on artificial intelligence |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108345941A (en) * | 2017-01-23 | 2018-07-31 | 阿里巴巴集团控股有限公司 | A kind of parameter regulation means and device |
| CN108345941B (en) * | 2017-01-23 | 2022-01-18 | 阿里巴巴集团控股有限公司 | Parameter adjusting method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| US20170255879A1 (en) | 2017-09-07 |
| KR20170102411A (en) | 2017-09-11 |
| JP2017157191A (en) | 2017-09-07 |
| JP6333342B2 (en) | 2018-05-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105631052A (en) | Artificial intelligence based retrieval method and artificial intelligence based retrieval device | |
| US10282462B2 (en) | Systems, method, and non-transitory computer-readable storage media for multi-modal product classification | |
| CN110263244B (en) | Content recommendation method, device, storage medium and computer equipment | |
| US11042898B2 (en) | Clickstream purchase prediction using Hidden Markov Models | |
| CN102314450B (en) | Method for enhancing webpage browse and equipment | |
| WO2021027256A1 (en) | Method and apparatus for processing interactive sequence data | |
| CN110737778A (en) | Knowledge graph and Transformer based patent recommendation method | |
| CN111242729A (en) | Serialization recommendation method based on long-term and short-term interests | |
| EP3155540A1 (en) | Modeling interestingness with deep neural networks | |
| CN110851699A (en) | Deep reinforcement learning-based information flow recommendation method, device, equipment and medium | |
| CN107463701A (en) | Method and apparatus based on artificial intelligence pushed information stream | |
| US20190354796A1 (en) | Systems and Methods for Slate Optimization with Recurrent Neural Networks | |
| US10762902B2 (en) | Method and apparatus for synthesizing adaptive data visualizations | |
| Shiue et al. | Development of machine learning‐based real time scheduling systems: using ensemble based on wrapper feature selection approach | |
| KR102446649B1 (en) | Method and apparatus for producing application using recommended templet based on artificial intelligence model | |
| KR20210012730A (en) | Learning method of artificial intelligence model and electronic apparatus | |
| Wang et al. | Intelligent dynamic control of stochastic economic lot scheduling by agent-based reinforcement learning | |
| CN112699310A (en) | Cold start cross-domain hybrid recommendation method and system based on deep neural network | |
| CN105243563A (en) | Commodity price predicting method and apparatus | |
| JP2024515373A (en) | Artificial neural network device and method for automatic design generation using UX-BIT | |
| CN114444807A (en) | Enterprise performance prediction method, system and storage medium | |
| US10146876B2 (en) | Predicting real-time change in organic search ranking of a website | |
| CN116910345A (en) | Label recommending method, device, equipment and storage medium | |
| CN115269984A (en) | Professional information recommendation method and system | |
| CN110020195A (en) | Article recommended method and device, storage medium, electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |